Go Back  PPRuNe Forums > Flight Deck Forums > Tech Log
Reload this Page >

AF 447 Thread No. 7

Tech Log The very best in practical technical discussion on the web

AF 447 Thread No. 7

Old 28th Dec 2011, 20:50
  #741 (permalink)  
 
Join Date: Jun 2009
Location: somewhere
Posts: 451
Yaw alternate and Dutch Roll:

Here some 'food for thought' considering Yaw alternate law and dutch roll dampening, extracted from AMM.

AMM 27-90 - ELECTRICAL FLIGHT CONTROL SYSTEM (EFCS)

(1) Normal configuration
----
----

(2) Reconfigurations
(a) Laws and associated functions
1 If the normal control laws are lost:
When the conditions required for keeping the normal control laws are no longer fulfilled, the control laws are reconfigured.
The various degraded law states that are possible are (in flight or upon flare):
// Roll and yaw:
Yaw alternate law
// Pitch:
Nz law (with limited pitch rate and gains)
Vc PROT law (ALPHA 2)
VMO2 law
Pitch direct law

The "Alternate" laws are engaged when the protections related
to the normal laws (ALPHA 1, VM01) are lost.
The "Direct" laws are engaged when the Nz law is lost.

The associated functions available are :
----
----
sideslip estimation (except in alternate 2 law or direct law)
if yaw alternate law, sideslip estimation is changed in Ny body lateral accelerometer.
----
----
2 Yaw alternate law

This FCPC and FCSC law is engaged if the lateral normal law is lost.
Its characteristics are:
- The roll control is direct, an order on the side stick directly commands a deflection, according to a kinematic calculation.
- The yaw control is achieved by the summation of two terms:
. pedal orders
. Dutch roll damping orders (from yaw rate)
In the event of loss of the inertial data from the ADIRUs, the yaw rate data for Dutch roll damping is provided to the FCPC via the rate gyro unit.
If the three FCPCs are lost, the FCSC1 ensures Dutch roll damping, using yaw rate data from the rate gyro unit.

(6) Turbulence Damping Function

(a) General
The purpose of the Turbulence Damping Function implemented in the Electrical Flight Control System is to damp the structural modes induced by atmospheric turbulence.

(b) Architecture
The Turbulence Damping Function consists of two lanes:
1 Longitudinal lane
The longitudinal Turbulence Damping command is computed by the FCPC1 (FCPC2 as a redundancy) as a function of the Nz accelerometer information.
It is added to the normal law command and transmitted to the associated elevator servo-controls.
2 Rear lateral lane
The rear lateral Turbulence Damping command is computed by the FCPC1 (FCPC2 and FCPC3 as a redundancy) as a function of the information of a specific Ny accelerometer located at the rear bulkhead level.
It is added to the normal law command and transmitted to the associated rudder servocontrols.
(c) Specific equipment
- the TURB. DAMP pushbutton switch
- the Ny rear accelerometer.
----
----
(11) Sideslip estimation and Beta target computation

(a) Sideslip estimation
This function is elaborated in the FCPCs in the flight and flare phases. The estimated sideslip is used as an input parameter for the lateral normal law. It is fed to the FMGECs and sent to the DMCs by the FCDCs for display on the PFD.
~o~

Here's a simplified schematic for the 'mechanical' rudder:

A33Zab is offline  
Old 16th Jan 2012, 20:55
  #742 (permalink)  
 
Join Date: Jun 2011
Location: france
Posts: 759
Snoop

Happy new Year to Everybody !

Thank You A33Zab !
Could we have a look inside S1, S2, P1, P3 ? ?
roulishollandais is offline  
Old 16th Jan 2012, 21:31
  #743 (permalink)  
 
Join Date: Jun 2011
Location: france
Posts: 759
Snoop

Is still anybody here ???

@A33Zab
Please, Do You have the answer to Grity's question (AF447 THREAD 7 post #730)

"what did you think is the frequency for the natural roll stability for a330 at this high and speed? "

In my own experience the frequency of the resonance was around 0.4 hz

@MachinBird
The Dutch roll is of much interest : equations have the lowest degree ! That means too that no additional energy is put in the closed loop (very very important).
For any other oscillations, you need to enlower this degree ! Otherwise you have not only inflexion points but inversion points in the response, and that cannot be managed by the pilot in flight.

Last edited by roulishollandais; 22nd Jan 2012 at 20:08. Reason: spelling MachinBird
roulishollandais is offline  
Old 17th Jan 2012, 01:15
  #744 (permalink)  
 
Join Date: Jun 2009
Location: somewhere
Posts: 451
@roulishollandais:

Could we have a look inside S1, S2, P1, P3
I guess a lot of electronic components gathered closely together....

We know what is going in and what is coming out. (FDR traces in--> Yaw damp order + pedal order
and out --> Rudder position --> Yaw rate = Yaw damp order....

If the lateral accel. trace is coming from ADIRU(below cockpit floor) the trace of Ny accelerometers
is missing (Ny accelerometers located aft of bulkhead).
AMM text:
"if yaw alternate law, sideslip estimation is changed in Ny body lateral accelerometer."
If this is the trace shown on the lateral parameter page it doesn't look very well....(after alternate law is engaged)

AMM text:
The estimated sideslip is used as an input parameter for the lateral normal law
Not mentioned, if the estimated slideslip is also a input during alternate operation or only used for display on PFD?
A33Zab is offline  
Old 4th Feb 2012, 16:41
  #745 (permalink)  
 
Join Date: Jun 2011
Location: france
Posts: 759
roulishollandais:
Quote:
inside S1, S2, P1, P3

A33Zab :
Quote:

We know what is going in and what is coming out.

Sure but I want to know the algorithms and logic circuits who are used, and time period of sampling , not only input and output ! I want to see INSIDE
roulishollandais is offline  
Old 4th Feb 2012, 19:59
  #746 (permalink)  
 
Join Date: Jan 2005
Location: France
Posts: 2,316
You want to look inside? Here you are...



What you really want is ...
- a full CMM (component maintenance manual),
- a full system block diagram,
- enough system engineering knowledge to interpret the above.....

PS..... OK, that's a Concorde AFCS circuit board.
In those days, the "algoritms" were 'frozen' in op-amps, resistors and capacitors, and the logic was hard-wired using TTL and DTL logic circuits.

The above photo is an analog control law computing board.

Here's where the logic is being dealt with...


Last edited by ChristiaanJ; 4th Feb 2012 at 20:42.
ChristiaanJ is offline  
Old 7th Feb 2012, 21:05
  #747 (permalink)  
 
Join Date: Jun 2011
Location: france
Posts: 759
Snoop analog and digital

Thank you very much ChristiaanJ ; this Concorde AFCS circuit board (analog control law computing board) is very beautiful; like the Concorde is ! How beautiful it was in the sky !

effectively what I still want is ...
- a full CMM (component maintenance manual),
- a full system block diagram,

Despite digital systems overcame on analog system, I am sure that the latter are the best to modelize and control non-linear unstable circuits.
Digitalizing brings many problems at the same time it seemed anything became easy by linearisation, and matrix computing.

Future is ahead !
roulishollandais is offline  
Old 8th Feb 2012, 20:50
  #748 (permalink)  
 
Join Date: Jun 2009
Location: somewhere
Posts: 451
@roulishollandais:

effectively what I still want is ...
- a full CMM (component maintenance manual),
- a full system block diagram
For sure a CMM will not supply you with the requested information.




For test and fail isolation the computer will be connected to an
ATEC Series 6 testbench.

The computer is a modular construction like any modern black box, if a card failed the card is replaced not the failed component on the card itself.

For obvious reasons (e.g. Intellectual property, trade secret, terrorism) a full logical diagram is hidden for the public,
but PJ2 posted a pitch channel logic for A320 a while ago:




FWIW you will not find beancounters in flight operations only, they are well spread over the world and also in the MRO business.
The suppliers of such sophisticated equipment are negotiating - aircraft life term - maintenance contracts (starting with B787)
for their equipment so all technology remain within the vendors company itself.
A33Zab is offline  
Old 9th Feb 2012, 20:32
  #749 (permalink)  
 
Join Date: Jan 2005
Location: France
Posts: 2,316
Originally Posted by A33Zab View Post
For sure a CMM will not supply you with the requested information.
If so, CMMs have changed from "my days".
A CMM would provide me with block diagrams, circuit diagrams, circuit board layouts, and yes, pictures like the one you showed, to correlate circuit board references and their physical location in the computer.
But it would also provide me with a full test specification.

For test and fail isolation the computer will be connected to an
ATEC Series 6 testbench.
I'm afraid you're confusing matters.
The ATEC is an automatic test system that runs an automatic test prgram.
That program does not just appear 'ex nihilo'... it is written on the basis of the original test specification, which can equally well be executed on a manual test bench or a different automatic test system (been there, done that, written ATEC programs....).

The computer is a modular construction like any modern black box, if a card failed the card is replaced not the failed component on the card itself.
That of course depends on what level maintenance you're talking about. But for first/second-line maintenance I agree.

For obvious reasons (e.g. Intellectual property, trade secret, terrorism) a full logical diagram is hidden for the public,
Depends on what you call "the public". In an avionics workshop, I would expect to have access to this kind of information.
....but PJ2 posted a pitch channel logic for A320 a while ago...
Terminology has obviously gone out of the window since "my days". That's not a 'logic' block diagram, but a 'functional' diagram.... and near-useless (except as an intro) to anybody called to maintain the system.
ChristiaanJ is offline  
Old 10th Feb 2012, 23:33
  #750 (permalink)  
 
Join Date: Jun 2009
Location: somewhere
Posts: 451
@ChristiaanJ:

Oke, we have an insider....

If so, CMMs have changed from "my days".
I think so, the specific FCSC CMM consists of 3 parts with a total of 2000+ pages.

Part A: General CMM (more ACMM with IPL and general description, diagrams, assembly etc) but only for casing and interconnecting board.
Part B: The Interconnections in detail
Part C: The ATLAS Test specification and it says - for ATEC series 6 -.

This ATLAS code says more to you than to me and the 'public'.
Anyway, for this type of computer, the required test equipment - today! - is the ATEC series 6.

The boards have their own CMM (I agree: for level 3 workshop) and the software logic and data resides in the OBRMs.

So, with this fragmented CMM information it will not help you to understand whats going on - inside -....
A33Zab is offline  
Old 10th Feb 2012, 23:54
  #751 (permalink)  
 
Join Date: Feb 2009
Location: Jungles of SW London
Age: 73
Posts: 354
ChristiaanJ

These are the boards you showed us on the "Concorde Question?" thread, I think, CJ? Nevertheless, they are good looking boards and bring back a lot of memories, but by gum they look dated now!

A33Zab

Thanks for posting those images - the top ones look just like my CT Renewal Parts manuals but the bottom one - the A320 Pitch channel logic - is 'positioned' rather oddly in my Medical Electronics experience. It is rather crude and a 'level' below most of the block diagrams I use, while being more than a level above the logic diagrams I used to use back in the day.

As CJ puts it:
Terminology has obviously gone out of the window since "my days". That's not a 'logic' block diagram, but a 'functional' diagram.... and near-useless (except as an intro) to anybody called to maintain the system.
You couldn't 'chase ones and zeros' through that and I wish I had a pound for every hour I've spent chasing them through page after page of logic diagrams.
Landroger is offline  
Old 23rd Feb 2012, 15:22
  #752 (permalink)  
 
Join Date: Jun 2011
Location: france
Posts: 759
testing the flying software and automation

@ChristiaanJ
#750
I totally agree with you.
The ATEC Series 6 testbench is a very bad choice who does the pilots blind.
The pilot must be able to understand very quickly what goes wrong in the flight to take the good decision.

The FCSC CMM with 2000 pages is a BABEL ATLAS.
roulishollandais is offline  
Old 23rd Feb 2012, 16:39
  #753 (permalink)  
 
Join Date: Jan 2005
Location: France
Posts: 2,316
Originally Posted by roulishollandais View Post
The ATEC Series 6 testbench is a very bad choice who does the pilots blind.
I don't quite get your point....

An ATEC is a huge computer-controlled automatic test bench, used for initial reception of the equipment, and for testing and fault-finding of equipment "thrown off the aircraft as suspect".
I think most pilots have never even seen one......

ATLAS is a specialised programming language for (mostly) avionics and similar aircraft equipment.
Describing ATLAS here in detail is really a bit too much O/T, but any old computer freaks will find some vague similarities with BASIC.
It really is a neat way of writing a full test specification for a piece of equipment in a format that can be understood directly by humans (and even performed on a manual test bench), as well as by fully automatic test equipment (such as the ATEC, but also ATEs from other manufacturers).

The pilot must be able to understand very quickly what goes wrong in the flight to take the good decision.
I agree, but that's not the same context....

The FCSC CMM with 2000 pages is a BABEL ATLAS.
What do you mean with that?
I believe the 2000 pages (ours were a bit smaller), but not all of that would be the ATLAS spec.... (unless the ATLAS progs for the individual boards were also included in the CMM, which I doubt), some would be intercon, wiring lists, circuit layouts, diagrams, IPCs, etc.

Anyway, all this is not too relevant, but a nice 'blast from the past'.
ChristiaanJ is offline  
Old 25th Feb 2012, 18:36
  #754 (permalink)  
 
Join Date: Jun 2011
Location: france
Posts: 759
fish

Hi ChristiaanJ,

[quote=ChristiaanJ]
Originally Posted by roulishollandais
Originally Posted by roulishollandais
The ATEC Series 6 testbench is a very bad choice who does the pilots blind.

I don't quite get your point....

An ATEC is a huge computer-controlled automatic test bench, used for initial reception of the equipment, and for testing and fault-finding of equipment "thrown off the aircraft as suspect".
I think most pilots have never even seen one......
Testing software or equipment is a very strategic action. ATEC is a good choice for EADS... If something is wrong or missing in ATEC it is the whole responsability if something gets wrong or missing in an EADS aircraft. But, seeing the ATEC advert. it seems to be an commercial product, that B could use too and B will then not be able to see something will get wrong or false in his aircraft or equipment...

Originally Posted by ChristiaanJ
I agree, but that's not the same context....
The flight engiener is no more in the aircraft. If some hydraulic or electric failure happens caution or warning light and call happen, and a check-list is used. If a software failure happens, nobody knows it in the airplane : the pilots are blinds, get unmotivated, and have behaviour like in AF447.
They get idiots.

Originally Posted by ChristiaanJ
2000 pages (ours were a bit smaller)
2000 pages of Basic : nice to write them, but too big to extract anything useful of such a FBW system for the pilot when things get crazy.

Before beeing a pilot, I had to organize computer math methods and software in a national research center. After that and due testing, the software was given to twelve regional centers. Some of these softwares concerned human life (dikes). They were running three years (you read well) at the same time that traditional method, without any difference in the results, and without changing anything in the software...
roulishollandais is offline  
Old 25th Feb 2012, 22:14
  #755 (permalink)  
 
Join Date: Jan 2005
Location: France
Posts: 2,316
roulishollandais,

We seem to be on slightly different wavelengths.....

A CMM is aimed at the maintenance engineer who gets a supposedly faulty piece of equipment dumped on his bench, and has to find and fix the supposed fault, and then recertify the equipment as airworthy.

It is not aimed at either a pilot or (in the olden days) at a flight engineer, who don't necessarily have the specialist engineering background to 'decipher' electronic circuit diagrams or test specifications in a document like the CMM.
They refer to the functional documentation (aircraft flight manuals, etc.).

Testing software or equipment is a very strategic action.
I think you're confusing testing during development and 'operational' testing....

The ATEC is strictly a production and 'operational' piece of test equipment. It has no relation with the equipment used for software and hardware testing and validation during design and development.
And indeed, ATECs have been sold to other clients, exactly like Honeywell, HP and SFENA ATEs.
ChristiaanJ is offline  
Old 25th Feb 2012, 23:52
  #756 (permalink)  
 
Join Date: Feb 2011
Location: Nearby SBBR and SDAM
Posts: 873
ATE's, Testability and Maintainability

Hi, ChristiaanJ


Automated and automatic testing was a passion during many years. I faced very interesting challenges designing for adequate testability at Module level (set of PCB's), Board level and Component level(LSI). Your comment on ATE made me ask you something:

Testability is a serious issue in a complex System as you know. A simple example may show why we need strategies to perform an effective test:

Suppose a VERY SMALL memory of just 64 bits. If you test it to all combinations of zeros and ones in the array at 10 Mhz the time required is prohibitive. The solution is to apply a checkerboard pattern, invert it, galloping ones, galloping zeros, all zeros, all ones, etc. And stil you may have a pattern sensitivity.

In a complex machine the Test Engineers use a multitude of Test Strategies at different phases (design phase, prototype testing, etc.)

My question is:

How you compare the ability to test thoroughly an A/C like a current FBW versus e.g. a Concord where Systems were less "Finite State Machines" equipped.

But highly complex and with "feedback loops" that creates formidable challenges when trying to locate for example, an intermittent failure.

I am always concerned with Complex Systems due the "testability issue", so this is the reason of my question to an Engineer from the Concorde era.

It comes to my mind something i heard on A330: To adjust the real Time Clock you have to lower the flaps a little bit.

Last edited by Jetdriver; 27th Feb 2012 at 08:03.
RR_NDB is offline  
Old 26th Feb 2012, 12:09
  #757 (permalink)  
 
Join Date: Jun 2009
Location: somewhere
Posts: 451
@RR_NDB:

It comes to my mind something i heard on A330: To adjust the real Time Clock you have to lower the flaps a little bit.
Well, nothing to do with a software bug...

Once a day (Flight Phase 9 = after landing) the wing tip brakes are automatically checked, if test fails to initialize for 10 days, this test should be performed by maintenance through MCDU.

When -incorrectly- setting the clock date (manual) you would go beyond those 10 days and trigger the FLAP TIP BRK FAULT and SLAT TIP BRK FAULT messages.
To reset these situation you have to reset the system through CMS AND cycle flaps opposite to the direction when it locked up.
If F/S are in the full up position you only have one way (=DOWN) to go for automatic reset. (and manual reset of the devices becomes necessary when a F/S Up command is required)

To prevent this situation you would lower the F/S a little bit so you can reset (by MCDU) in both directions.

With the GPS clock setting option this would not bring one into this trouble.
A33Zab is offline  
Old 26th Feb 2012, 21:36
  #758 (permalink)  
 
Join Date: Feb 2011
Location: Nearby SBBR and SDAM
Posts: 873
...you have to lower the flaps a little bit

Hi,

A33Zab

Well, nothing to do with a software bug...
The context of my post was Testability, maintainability, etc.

Once a day (Flight Phase 9 = after landing) the wing tip brakes are automatically checked, if test fails to initialize for 10 days, this test should be performed by maintenance through MCDU.

When -incorrectly- setting the clock date (manual) you would go beyond those 10 days and trigger the FLAP TIP BRK FAULT and SLAT TIP BRK FAULT messages.
To reset these situation you have to reset the system through CMS AND cycle flaps opposite to the direction when it locked up.
If F/S are in the full up position you only have one way (=DOWN) to go for automatic reset. (and manual reset of the devices becomes necessary when a F/S Up command is required)

To prevent this situation you would lower the F/S a little bit so you can reset (by MCDU) in both directions.

With the GPS clock setting option this would not bring one into this trouble.
Fine biz! Some friends also will benefit from your explanation.

Rgds,
RR_NDB is offline  
Old 26th Feb 2012, 22:06
  #759 (permalink)  
 
Join Date: Feb 2011
Location: Nearby SBBR and SDAM
Posts: 873
RHS not recorded

Hi,

A33Zab,

Is this a concern in order to understand reasons of PF persistent NU?

Mismatch between sides are recorded?

BEA published information mentioned this?

If existed a mismatch how we could learn what PF saw?

If this specific point was covered earlier, please inform the link, if possible.
RR_NDB is offline  
Old 27th Feb 2012, 07:57
  #760 (permalink)  
 
Join Date: Jul 2009
Location: France - mostly
Age: 80
Posts: 1,689
Originally Posted by RR NDB
BEA published information mentioned this?
See for example BEA Interim #3(en) p.44 "1.16.6 Reconstruction of the information available to the crew".
HazelNuts39 is offline  

Thread Tools
Search this Thread

Contact Us - Archive - Advertising - Cookie Policy - Privacy Statement - Terms of Service - Do Not Sell My Personal Information

Copyright 2018 MH Sub I, LLC dba Internet Brands. All rights reserved. Use of this site indicates your consent to the Terms of Use.